Note: Descriptions are shown in the official language in which they were submitted.
CA 02351966 2001-06-29
Moeller 9-12
METHOD AND APPARATUS FOR POLARIZATION
MEASUREMENTS
C'R(W~ RFFFRFT'TC'FS TO RFT ATFT1 hDDT T~rmTn,,rc~
This application claims priority to U.S.
provisional application 60/228,265, entitled
"Method and Apparatus for Polarization
Measurements," filed August 25, 2000, which is
incorporated herein by reference in its entirety.
This application is related to U.S.
provisional application 60/229,129, entitled
"System Comprising In-Line Wavelength Sensitive
Polarimeter," filed August 30, 2000, which is
incorporated herein by reference in its entirety.
TFCHNTCA1, FTFT 17
The invention relates to a method and
apparatus for polarization measurements and, more
particularly, to a method and apparatus for
frequency-resolved polarization measurements of
an optical signal in wavelength division
multiplexed systems.
BAC'KCROTTNT) OF THF T TVFTyTTTIInT
The demand for transmission capacity
enhancement in fiber networks and for low cost
per bit for high bit-rate systems requires
continuous improvement in single channel bit-
rates. Electrical time division multiplexed
(ETDM) systems at bit-rates of 40 Gb/second per
channel are commercially available, and optical
time division multiplexed (OTDM) systems offering
bit-rates of 160 Gb/s per channel have been
demonstrated.
CA 02351966 2001-06-29
Moeller 9-12
Aside from other transmission impairments
like chromatic dispersion or non-linear pulse
propagation, polarization mode dispersion (PMD)
remains an obstacle to be overcome before these
ultra high time division multiplexed (TDM)
systems can be implemented on a significant
number of installed fibers. Several different
PMD compensation and mitigation techniques have
been proposed and partially demonstrated. They
can be generally categorized according to
concepts based on the electrical or optical
domains of the system. Electrical mitigation
techniques possess advantages-in terms of cost
and size. However, the performance improvements
from electrical techniques are generally not as
great as those achievable with optical
compensation techniques.
For example, optical techniques can be
designed to be bit-rate and format independent.
The measurement of the polarized components of an
optical signal may be used for implementing PMD
compensation using optical techniques.
Aside from PMD compensation, simple
polarization monitoring will also be an important
feature in future networks employing polarization
multiplexing. Furthermore, polarization
measurement is required as well for signal
polarization adaptation to polarization sensitive
receivers.
~ITMMARY OF THF TN 7FtTTTn 1
The invention relates generally to a method
and apparatus for polarization measurements. A
polarization state of an optical signal can be
CA 02351966 2001-06-29
Moeller 9-12
determined using a polarization analyzer
comprising a polarization controller, a
polarizer, a wavelength dispersive element, and a
photo-detector.
In one embodiment, the photo-detector is a
photodiode array. To determine the polarization
state of the optical signal, optical power
measurements are performed as a function of
wavelengths within the optical signal for
different settings of the polarization
controller. The method can be applied to
polarization and polarization mode dispersion
determinations -in wavelength division-multiple~ed
(WDM) systems.
BRT_RF 17RSC'RTPTTnt~T n~ THF DRAWTI~Trc
The teachings of the present invention can be
readily understood by considering the following
detailed description in conjunction with the
accompanying drawings, in which:
FIG. 1 is a schematic representation of a
conventional polarization analyzer;
FIG. 2 is a schematic representation of a
polarization analyzer according to one embodiment
of the invention;
FIGS. 3a-c are schematic illustrations of
multichannel detection of a dispersed optical
signal;
FIG. 4 is a flow diagram of a method of
polarization measurement according to one
embodiment of the invention;
FIG. 5 is a schematic representation of a
transmission system incorporating one embodiment
of the invention;
CA 02351966 2001-06-29
Moeller 9-12
FIG. 6 is a schematic illustration of one
example of a polarization analyzer of the
invention; and
FIG. 7 is a schematic illustrat~.on of another
example of a polarization analyzer of the
invention.
To facilitate understanding, identical
reference numerals have been used, where
possible, to designate identical elements that
are common to the figures.
T7FTATT FT7 D ,SC'RTPTTnN nF THR TNVRNTTON
The invention provides generally a method and
an apparatus for measuring a polarization state
of an optical signal. Polarization measurements
1~ can be performed using a polarization analyzer
(or polarimeter) comprising a polarization
controller, a polarizer, a wavelength dispersive
element, and a photo-detector. In one
embodiment, the polarization state and the power
of a broadband optical signal are obtained by
measuring the Stokes vector components of the
optical signal.
The method and apparatus can also be applied
to polarization mode dispersion (PMD)
measurements. For example, embodiments of the
invention can be applied to on-line measurements
of polarization and PMD in a fiber link of a wave
division multiplexed (WDM) system without data
traffic interruption. By analyzing the frequency
dependence of the polarization of the optical
signal at a receiver input, the PMD
characteristic of a WDM channel for the
corresponding frequency interval can be
CA 02351966 2001-06-29
Moeller 9-12 5
determined. Results of the PMD measurements can
be used to derive PMD compensation parameters.
Alternatively, PMD compensation can also be
implemented by measuring the degree of
polarization (DOP) of a WDM signal using the
polarization analyzer. Although the invention
can be practiced in conjunction with any type of
PMD compensator, it is particularly well-suited
for use with optical compensators.
FIG. 1 is a schematic representation of a
conventional polarimeter 100, e.g., a model HP
8509A, available from Hewlett-Packard, of Palo
Alto, California. Details of such a polarimeter
100 are disclosed in U.S. patent 5,227,623,
entitled "Method and Apparatus for Measuring
Polarization Mode Dispersion in Optical Devices,"
which is incorporated herein by reference in its
entirety.
An optical signal 150 having a polarization
state to be determined is coupled to the
polarimeter 100 via an optical connector 102.
The optical signal 150 is directed onto a
spherical mirror 104 and split into four
components A, B, C and D by four adjustable
quadrants of the spherical mirror 104. As shown
in FIG. 1, three of the four components (A, B and
C) are coupled into respective polarizers P1, P2
and P3. Prior to entering the polarizer P1, one
of the optical signal components, e.g., A, is
passed through a waveplate 106. The waveplate
106 is used to invert the polarization state of
the signal component A. Each polarizer P1, P2
and P3 is set to different orientations, such
CA 02351966 2001-06-29
Moeller 9-12
that after passing through the respective
polarizers, the optical components, denoted by
A', B' and C', are characterized by different
polarization components.
The intensities or powers of the four optical
signal components A', B', C' and D are measured
using photodiodes PD1, PD2, PD3 and PD4. The
power of the fourth component, D., which is not
filtered by any polarizer, is measured directly
by the photodiode PD4. The photodiodes PD1, PD2,
PD3 and PD4 provide four output photo-currents
I1, I2, I3 and I4, which are proportional to the
measured powers of the respective optical signal
components A', B', C' and D. The Stokes
1S parameters, which characterize the polarization
state of the input optical signal 150, can be
determined by analyzing the photo-currents I1,
I2, I3 and I4, using techniques that are known in
the art.
FIG. 2 is a schematic representation of a
polarization analyzer 200 according to one
embodiment of the invention. The polarization
analyzer 200 comprises a polarization controller
202, a polarizer 204, a wavelength dispersive
element 206, and a photo-detector 208.
The polarization controller 202 is used to
perform defined polarization transformations of
an incoming optical signal 250. Different types
of polarization controllers 202 may be used to
practice various embodiments of the invention,
including mechanical, electro-mechanical and
electro-optical devices. In general, mechanical
devices tend to be slow compared to electro-
CA 02351966 2001-06-29
Moeller 9-12
optical ones. Therefore, electro-optical
polarization controllers, e.g., lithium niobate-
based devices, are preferable because they allow
polarization state transformation to be performed
at relatively high speeds, e.g., on the order of
MHz.
In one embodiment, the polarizer 204 is a
linear polarizer, and is selected to operate
within a wavelength region of the optical signal
250.
The wavelength dispersive element 206 is used
to disperse, or separate, the optical signal 250
into its spectral components within a wavelength
range f rom min t~ "'max ~ where ~~min and a,max denote
IS the lower and upper wavelength limits for the
spectral range of the optical signal 250.
Depending on their wavelengths, different
spectral components are dispersed into different
angles. It should be noted that the wavelengths
7~; and ~,~ are used to denote spectral components
characterized by central wavelengths ~.i and ~,~ ,
though each spectral component may further be
characterized by a spectral range around the
central'wavelengths.
The polarization analyzer 200 is generally
applicable to polarization measurements of input
optical signals having either monochromatic or
broadband spectra. For WDM applications, the
optical signal 250 typically has a spectrum
residing in telecommunication windows around
1.3~m, 1.5~tm and 800nm regions. In principle,
CA 02351966 2001-06-29
Moeller 9-12
the polarization analyzer 200 can be adapted to
these signal wavelength regions.
The wavelength dispersive element 206 should
have a bandwidth that is sufficiently large to
encompass the bandwidth of the optical signal
250. In one embodiment, the wavelength
dispersive element 206 is a grating, e.g., a
diffraction grating, with a bandwidth of about
32 nm and a spectral resolution of about 0.1 nm.
Other types of wavelength dispersive element 206
may also be used, e.g., a fiber grating, a prism
or scanning etalon. In general, it is preferable
- that the wavelength dispersive element 206 does
not involve mechanical scanning, in order to have
improved reliability and increased speed for
polarization measurement of a broadband signal.
Referring to FIG. 2, the dispersed spectral
components (~i, ~~, ...) of the optical signal
250, which are separated spatially as a function
of wavelength, are detected by a photo-detector
208. In one embodiment, the photo-detector 208
is a photodiode array, which comprises a series
of pixels for detecting the dispersed spectral
components (~i, ~~, ...) of the optical signal
250. One example of the photodiode array 208 is
an optical channel monitor, which can be used for
wavelength, power and optical signal to noise
ratio (OSNR) measurements. Details of an optical
channel monitor have been disclosed in U.S.
Patent 6,078,709, entitled "Method and Apparatus
for Monitoring Multi-Wavelength Optical Systems,"
issued on June 20, 2000, and in U.S. Patent
6,002,822, entitled "Article Comprising An
CA 02351966 2001-06-29
Moeller 9-12 g
Optional Waveguide Tap," issued on December 14,
1999, both of which are incorporated herein by
reference. It is understood that other
photodiode arrays or optical channel monitors may
also be suitable for practicing embodiments of
the invention. If the polarization analyzer 200
is used to measure polarization of a
monochromatic optical signal, then a single
photo-detector or photodiode may also suffice.
For WDM applications, spectral resolution of
the polarization analyzer 200 is an important
design parameter. The spectral resolution of the
polarization analyzer 200 is determined by the
combined resolution of the wavelength dispersive
element 206 and that of the photodiode array 208.
In order to perform frequency-sensitive (or
wavelength-dependent) polarization measurements
corresponding to different spectral components,
the resolution of the polarization analyzer 200
should be sufficiently high to ensure that the
resolved frequency interval is small enough such
that the Stokes vector remains substantially
constant within the resolution interval. In
general, the appropriate resolution depends on
the degree of accuracy that is desired for the
polarization measurements, and may vary according
to the specific needs or applications. It is
understood that one skilled in the art would
select a resolution to provide measurements
within acceptable limits for specific
applications.
FIG. 3a illustrates schematically the
detection of the dispersed spectral components
CA 02351966 2001-06-29
Moeller 9-12 10
(e.g. , ~,;, 7~;, 7~k, . . . ) by different pixels or
channels (PXi, PX~, PXk, . . . ) of the photodiode
array 208. With sufficiently high resolution,
each pixel of the detector array 208 detects a
portion of the spectrum of the optical signal
250, corresponding to a spectral component that
is sufficiently narrow such that there is no
significant change in the Stokes vector, or
polarization. As illustrated in FIG. 3a,
spectral components ~,i, ~~, ~,k (each with its
associated spectral range ~~,;, 0~,~, 47~k) , are
detected respectively by adjacent-pixels PX;, PX~,
and PXk. If the Stokes vectors for the spectral
components ?~=, ~,~, 7~k remain substantially constant
within their respective spectral ranges, the
optical powers detected by pixels PXi, PX~, and
PXk will correlate with the polarization
parameters for the respective spectral
components.
In contrast, FIG. 3b illustrates
schematically the detection of dispersed spectral
components with insufficient resolution of the
polarization analyzer 200. As shown in FIG. 3b,
one pixel 301 of the photodiode array 208
simultaneously detects several spectral
components ~.;, ~,~, ~,k, each with slightly different
polarization vectors Si, S~, Sk. The intensity
(or contribution) of each spectral component, as
detected by the pixel 301, is determined by the
filter function of the wavelength dispersive
element 206. Thus, instead of detecting signal
contribution from one polarization state
CA 02351966 2001-06-29
Moeller 9-12 11
corresponding to a specific spectral component,
the pixel 301 measures an average polarization
state with a degree of polarization (DOP) smaller
than one. The DOP describes how polarization
vectors of different spectral components are
aligned with respect to each other within the
wavelength resolution of one pixel. If the
polarization vectors are completely aligned, then
DOP is at a maximum value of one. However, if
the polarization vectors are not completely
aligned, then DOP is less than one.
FIG. 3c illustrates yet another scenario in
which each of the spectral components ~~i, 7~~ is
detected by several detector pixels of the
IS photodiode array 208. In this example, the
spectral component ~,i is detected by three pixels
303, 305 and 307; while the spectral component
is detected by three pixels 307, 309 and 311.
Since pixels 303, 305, 309 and 311 contain signal
contribution from only one spectral component ~,;
or 7~~, the photo-currents from these pixels (e. g.,
sum of 303 and 305 and sum of 309 and 311) are
proportional to the signal intensity for the
individual spectral components ~,; and ~,
Although the photo-current from pixel 307
contains contributions from both spectral
components ~,; and 7~~, it may still be possible to
extract, using appropriate signal processing
techniques, individual signal contributions for
the determination of the polarization states for
the spectral components ~.i and
CA 02351966 2001-06-29
Moeller 9-12 12
In general, for WDM applications, the
polarization analyzer 200 should have sufficient
resolution to avoid polarization averaging over a
wavelength interval in which the polarization or
S optical power can significantly vary. For
example, it is preferable to avoid polarization
averaging within one pixel, such as that
illustrated in FIG. 3b, or at least within the
signal bandwidths of adjacent WDM channels.
Methn~ ~f Pnlari~ati~n M aim rPmanr
The polarization state of an optical signal
can be determined using the polarization analyzer
200 of FIG. 2 by measuring the powers of the
optical signal corresponding to several different
polarization transformations of the polarization
controller 202.
FIG. 4 is a flow diagram of a method of
polarization measurement according to one
embodiment. In this embodiment, the Stokes
vectors for the spectral components of the
optical signal are measured and the polarization
state and optical power of an optical signal can
both be determined from the Stokes components.
The method can generally be applied to a
variety of optical signals with different
spectral ranges. For example, the optical signal
may have a wavelength range in the 1.55 ~m
region, such as one that is used in WDM systems.
Alternatively, the optical signal may have a
broadband spectrum in other wavelength regions.
Furthermore, the method may also be applied to a
monochromatic optical signal.
CA 02351966 2001-06-29
Moeller 9-12 13
At step 401, an input optical signal,
characterized by an arbitrary polarization state
to be determined, is launched into the
polarization analyzer 200. At step 403, the
polarization controller 202 is set to a first
position, and the optical signal is converted
from its arbitrary polarization state to a
transformed polarization state PC1.
At step 405, the optical signal is directed
to the linear polarizer 204, which is set at a
fixed orientation to define a polarization axis
for the exiting optical signal.
At step 407, the optical signal exiting the
polarizer 204 is directed onto a wavelength
1~ dispersive element 206, which disperses the
spectrum of the optical signal 250 as a function
of wavelength. In general, the dispersed
spectrum of the optical signal 250 may correspond
to a continuum of wavelengths, or to a spectrum
consisting of wavelengths corresponding to
separate communication channels. In either case,
the dispersed spectrum can be considered as being
made up of different spectral components, with
each spectral component characterized by a
central wavelength and an associated spectral
range (or spectral width). In the case of a WDM
optical signal, one or more of the spectral
components may correspond to individual optical
channels of the WDM system.
At step 409, the optical powers of the
spectral components are measured by a photo-
detector 208 such as a photodiode array. Each of
the spatially separated spectral components may
CA 02351966 2001-06-29
Moeller 9-12 14
be directed to one or more pixels of the
photodiode array 208. A photo-current, which is
proportional to the optical power of the detected
spectral component, is generated in the
photodiode array 208. The magnitudes of the
measured photo-currents for the respective
spectral components of the optical signal may be
stored in a memory or processor for calculation
of the Stokes components of the optical signal.
There are two aspects in practicing the
invention for WDM applications. One aspect
involves measurements of wavelength-dependent
polarization (e. g., for a signal with a non-zero
PMD), in which case, each signal corresponding to
a WDM channel, which may comprise several
spectral components, illuminates (i.e., be
detected by) several pixels. However, at least
some pixels of the photodiode array should each
detect primarily a frequency or wavelength range
within which the Stokes vector remains
substantially constant, in order to avoid
polarization averaging.
The other aspect involves measuring the
degree of polarization (DOP) of the optical
signal. In this case, it is sufficient that each
signal corresponding to a WDM channel illuminates
only one pixel in the photodiode array (even if
the polarization state changes within the
bandwidth of one WDM channel), as long as no two
signals corresponding to two WDM channels
illuminate the same pixel. However, the
resolution of the polarization analyzer, which
can be limited by the wavelength dispersive
CA 02351966 2001-06-29
Moeller 9-12 15
element, should at least be sufficiently high to
avoid polarization averaging between different
WDM channels. In either aspect of practicing the
invention, the resolution is selected to provide
measurements within acceptable tolerance for the
specific applications.
Similar optical power measurements are
. subsequently performed, as shown in step 411,
with the polarization controller 202 set to at
least three other positions, each of which is
different from the first position and from each
other. In general, many different combinations
of four positions may be used for the
polarization controller 202 in performing the
polarization measurement according to the method
of the invention. For example, one possible
combination of four settings of the polarization
controller 202 may consist of launching a signal
with transverse electric (TE) mode into the
polarization controller 202. The polarization
controller 202 may be set sequentially to produce
four output or transformed polarization states.
If one uses a Poincare sphere representation of
Stokes vectors for describing the polarization
2S states of the signal, then these four transformed
polarization states should preferably have
maximum spacings from each other on the Poincare
sphere -- e.g., forming a tetrahedron.
For each of the spectral components, the
measured optical powers corresponding to the four
positions of the polarization controller 202 may
be analyzed, using relationships that are known
in the art, to generate or calculate the Stokes
CA 02351966 2001-06-29
Moeller 9-12 16
vector components, or Stokes parameters, which
describe polarization and optical power of the
optical signal.
In general, the four power measurements are
related to the four Stokes parameters by a 4x4
matrix, known as a calibration or instrument
matrix, which may be determined by a calibration
procedure known in the art. Details of a
suitable calibration procedure has been described
by R.M.A. Azzam, in "In-line, Light-Saving
Photopolarimeter and its Fiber-Optic Analog",
Optics Letters, Vol. 12, No. 8, pp.558-560, 1987,
which is incorporated herein by reference in its
entirety. Depending on the wavelength dispersive
element, or optical couplings between the
polarization controller, the polarizer and the
wavelength dispersive element, the calibration
matrix may potentially be different for every
detector pixel. Once the calibration matrix has
been determined, the four Stokes parameters can
be obtained by multiplying the four power
measurements generated by each detector pixel
with the corresponding calibration matrix for the
detector pixel.
The polarization state corresponding to the
spectral components may then be determined from
the Stokes parameters. It is noted that power
measurements corresponding to at least four
transformed polarization states PC1, PC2, PC3 and
PC4 (corresponding to four positions of the
polarization controller 202) are necessary to
provide sufficient data to determine the original
polarization state described by the Stokes vector
CA 02351966 2001-06-29
Moeller 9-12
and the power of the input optical signal.
Additional measurements, i.e., corresponding to
more than four transformed polarization states,
are not precluded, although they are not
necessary.
It is understood that the method of FIG. 4
illustrates only one application of the
polarization analyzer 200. Thus, the
polarization analyzer 200 may also be used in
other applications which do not require
measurements corresponding to four different
settings of the polarization controller 202. For
example, if the power of the input optical signal
is not required, then less than four measurements
may suffice to provide polarization or other
desired information for the optical signal.
Thus, depending on the specific information
desired, the polarization analyzer 200 can
generally be used for various optical
measurements by generating different numbers of
transformations corresponding to different
positions (may be more than or less than four) of
the polarization controller 202.
Several examples are provided below to
further illustrate different polarization
transformations that can be used for practicing
the invention.
Rxam 1~ P l
FIG. 6 is a schematic illustration of a
polarization analyzer 600, comprising a
polarization controller 602, a polarizer 608, a
wavelength dispersive device 610, and a
photodetector array 612. In this example, the
CA 02351966 2001-06-29
Moeller 9-12 Ig
polarization controller 602 consists of a half-
wave plate 604 and a quarter-wave plate 606. An
optical signal 650 having a certain polarization
state passes through the half-wave plate 604, the
quarter-wave plate 606 and the polarizer 608.
After exiting the polarizer 608, the optical
signal 650 is dispersed by the wavelength
dispersive device 610, and the dispersed signal
is subsequently detected by the photodetector
array 612.
The orientations of the half-wave plate 604,
the quarter-wave plate 606 and the polarizer 608
are repT~esented respectively by 812 , 61~~ and 8a .
Table 1 provides one example of the four
IS positions or transformations of the polarization
controller 602 (each position being defined by
and 8a) that can be used for practicing
the invention. The orientations of the half-wave
plate 604 and the quarter-wave plate 606 are
given as angular rotation (in degrees) relative
to the orientation of the polarizer 608. By
performing power measurements for the dispersed
signal at each of the four positions of the
polarization controller 602, Stokes parameters
for the optical signal 650 can be obtained and
used for the determination of the polarization
state of the optical signal.
T~hl P
e1~2 e1~4 ea
~0 1 0 0 0
2 22.5° 0 0
3 45° 0 0
CA 02351966 2001-06-29
Moeller 9-12 19
4 22.5° 45° 0
In another embodiment, the order of the half-
wave plate 604 and the quarter-wave plate 606 may
be interchanged such that the optical signal 650
passes through the quarter-wave plate 606 before
the half-wave plate 604. In that case, the
orientations of the half-wave plate 604 and the
quarter-wave plate 606 corresponding to the four
transformations of the polarization controller
602 will be different from those shown in Table
1. One skilled in the art will be able to devise
suitable orientations for practicing alternative
embodiments of the invention.
Rx~m
1~ FIG. 7 is a schematic illustration of
polarization analyzer 700, comprising a
polarization controller 702, a polarizes 704, a
wavelength dispersive device 706, and a
photodetector array 708. In this example, the
polarization controller 702 is a quarter-wave
plate with an orientation denoted by 914. To
practice the invention, an optical signal 750
having a certain polarization state passes
through the polarization controller 702 and the
polarizes 704. The signal 750 exiting the
polarizes 704 is dispersed using the wavelength
dispersive device 706, and the dispersed signal
is detected by the photodetector array 708.
With the polarizes 704 maintained at a fixed
orientation 6a, the orientation 614 of the
polarization controller 702 is varied
continuously as a function of time - e.g., by
rotating the quarter-wave plate through different
CA 02351966 2001-06-29
Moeller 9-12 20
positions at a given rate. The dispersed signal
intensity (I) detected by the photodetector array
708 can be represented by the following equation:
I - 0.5 [ (So + S1/2) + (S1/2) cos (4014)
+ (SZ/2) sin(461~4) - S3sin(201~4)
Eq. (1)
where So, S1, S2, and S3 are the respective Stokes
parameters of the optical signal 750.
The Stokes parameters So, S1, S2, and S3 may
be obtained by performing a Fourier analysis of
the detected signal spectrum. By performing
Fourier analysis for all spectral components, a -
wavelength dependent polarization of the optical
signal 750 can be determined. Techniques of
performing Fourier analysis for single wavelength
applications have been disclosed in "Survey of
Methods for the Complete Determination of a State
of Polarization", by P.S. Hauge, in SPIE vol. 88,
Polari .ec~ ,i~r, pp.3-10, 1976, which is
incorporated herein by reference in its entirety.
Rxam lPP
In another embodiment, the polarization
analyzer may be used to measure spectral power
density of an optical signal having a certain
polarization state. In this case, it is only
necessary that the polarization controller be set
at two different positions to provide the
transformations necessary for power density
measurements. In one example, the two positions
of the polarization controller may differ from
each other by 90°, producing transformations
corresponding to the TE and TM modes
respectively. Depending on the specific
CA 02351966 2001-06-29
Moeller 9-12 21
applications, different numbers of orientations
of the polarization controller may also be used
to generate other optical measurements, as
appropriate.
(7n-1 snP PMD MPa~mrPmPnt i n a Tran~mi s~i ~n ~~r~r~m
According to another aspect of the invention,
the polarization analyzer 200 can be used for
measurement of polarization mode dispersion (PMD)
in a transmission system. In particular, it can
be used as part of a PMD monitoring unit for on-
line PMD measurement, concurrent with data
transmission (i.e., without interrupting data
transmission) in a WDM system. On-line PP7D
measurements using a conventional polarization
1~ analyzer has previously been disclosed in a
commonly-assigned U.S. patent application, Serial
No. 09/518,296, entitled "Method and Apparatus
for On-Line Monitoring of Polarization Mode
Dispersion in a Transmission System," filed on
March 3, 2000, which is herein incorporated by
reference in its entirety.
FIG. 5 shows a schematic diagram of a WDM
transmission system 500 containing a PMD monitor
550 for. on-line PMD measurement. The PMD monitor
550 comprises a polarization switch 502 and the
polarization analyzer 200.
In the WDM transmission system 500, different
transmitters (TX1, TX2, :.., TXN) are used to
generate light (or optical carrier) at different
wavelengths for data transmission. The carrier
signals are modulated by respective data bit
streams to form data signals DS1, DS2, ..., DXN.
These data signals DSl, DS2, ..., DXN, having
CA 02351966 2001-06-29
Moeller 9-12 22
different carrier wavelengths are combined in a
multiplexer 510 to form a single optical signal
580.
The optical signal 580, which is
S illustratively, linearly polarized, is coupled
into a polarization switch 502 before entering the
transmission line 512. The polarization switch
502 allows the.output polarization of the optical
signal 580 to be varied. For example, the optical
signal 580 may be switched between two non-
orthogonal linear polarization states, which have
a relative angle of 45° (in Jones space) with _
respect to each other. Many devices may be used
as the polarization switch 502 for rotating the
polarization of the optical signal 580, with one
example being a Pockels cell.
In the schematic representation of FIG. 5, it
is assumed that the transmission fiber 512
represents the entire transmission line
characterized by a polarization mode dispersion
(PMD) to be monitored using embodiments of the
invention. At its output end 514, the
transmission fiber 512 is connected respectively
to the polarization analyzer 200 and a PMD
compensator 516. As such, the transmission signal
580 is split into two portions -- one portion 580a
being coupled into the polarization analyzer 200,
and another portion 580b being coupled to the PMD
compensator 516.
The polarization analyzer 200 is also
connected to a computer or a microprocessor 518
that communicates with the polarization analyzer
200 and a controller 520 for controlling the PMD
CA 02351966 2001-06-29
Moeller 9-12 23
compensator 516. Alternatively, the controller
520 may be incorporated as part of the computer
518.
The PMD compensator 516 is connected at its
output 522 to a demultiplexer 524 for separating
the transmitted signal 580b into its respective
channel components DS1, DS2, ..., and DSN. The
optical signals of the respective channels are
detected by a number of receivers RX1, RX2, ...,
RXN. Typically, the polarization analyzer 200 and
the PMD compensator 516 are located in close
proximity to the receivers RX1, RX2, ..., RXN, to
allow accurate PMD compensation for the data
signal 580b arriving at the receivers RX1, RX2,
..., RXN.
In an alternative embodiment, the
polarization analyzer 200 may be connected to the
output 522 of the PMD compensator 516 such that
the transmitted signal 580 exiting the PMD
compensator 516 may be split into two portions,
with one portion directed to the polarization
analyzer 200, and another portion to the
demultiplexer 524. This arrangement is
illustrated in phantom in FIG. 5. In general, the
polarization analyzer 200 of the invention can be
used in conjunction with any PMD compensator 516,
including electrically-based and optically-based
compensators.
The polarization switch 502 and the
polarization analyzer 200, which collectively
form the PMD monitor 550, can be used to provide
real-time, on-line PMD measurements for the
transmission fiber 512. Since the signal source
CA 02351966 2001-06-29
Moeller 9-12 24
used for PMD measurement is provided by the
transmission or data signal 580 (as opposed to an
external signal source), embodiments of the
invention allow PMD measurements to be performed
concurrent with data transmission in the
transmission system 500, without interrupting
data transmission.
To determine the PMD of the transmission
fiber 512, it is necessary to perform
polarization measurements for at least two
different and non-orthogonal polarization states
of the optical signal 580 launched into the
transmission fiber 512. For example, with the
polarization switch 502 set at a first
orientation, a first set of polarization
measurements may be performed according to steps
such as those outlined in FIG. 4. Subsequently,
a second set of polarization measurements is
performed with the polarization switch 502 set at
a second orientation, e.g., about 45° with respect
to the first position. It is noted that
polarization switching should not affect the
signal propagation in an intensity modulated
direct detection (IMDD) system, at least in the
absence of fiber nonlinearities, as long as
polarization dependent losses can be neglected,
and PMD has been compensated for. The measured
wavelength dependency of the Stokes vectors
(i.e., polarization measurement as a function of
wavelength) allows determination of PMD in the
transmission fiber 512.
Results of these measurements are provided as
input data to the computer or microprocessor 518,
CA 02351966 2001-06-29
Moeller 9-12 25
which then computes relevant parameters for the
adjustment of the PMD compensator 516.
Alternatively, based on results of the PMD
measurements, PMD compensation parameters may
also be retrieved from a look-up table that may
be stored in the computer 518. These parameters
are communicated to the controller 520 which then
adjusts the PMD compensator~516, such that the
optical signal 580b arriving at the demultiplexer
524 is compensated for any pulse distortion due
to PMD in the transmission system 500.
According to another aspect of the invention,
the polarization analyzer 200 may also be used to
measure the degree of polarization (DOP) for PMD
IS monitoring purpose - through measurement of the
Stokes vectors of the transmitted signal 580. For
example, the DOP may be used as an indicator of
the presence of PMD in the transmitted signal 580,
and such information may be used by the processor
518 to generate a control signal for controlling
the PMD compensator 516.
In one embodiment, the polarization analyzer
200 is connected to the output of the PMD
compensator 516, such that one portion of the
transmitted signal 580 enters the polarization
analyzer 200 and the other portion enters the
demultiplexer 524. The DOP may be measured using
the polarization analyzer 200 by performing
optical power measurements for individual WDM
channels, which are resolved by passing the
transmitted signal 580 (after propagating through
the polarization controller and the polarizer 202
and the polarizer 204) through the wavelength
CA 02351966 2001-06-29
Moeller 9-12 26
dispersive element. The optical powers measured
by the detector array 208 corresponding to each
of the individual WDM channels are then used to
calculate Stokes componEnts for the WDM channels,
which in turn allows a determination of the DOP.
The DOP data can be analyzed to derive
information about PMD, which can then be used for
controlling the PMD compensator 516.
Although several preferred embodiments which
incorporate the teachings of the present
invention have been shown and described in
detail, those skilled in the art can readily
devise many other varied embodiments that still
incorporate these teachings.
